Osteoporosis is an enormous public health problem causing at least 1.3 million fractures and costing up to $10 billion in the United States each year. Our specific aims attempt to fill gaps in knowledge that impede the design and implementation of an effective control program. 1) The age- stratified, random sample of 262 Rochester, MN women who have been followed for up to 8 yrs with serial measurements of bone mineral density (BMD) and having had extensive baseline studies fare a unique resource. To obtain long-term data on the patterns and causes of bone loss and fractures, we will extend their followup to 12 yrs. 2) Novel biochemical indices of bone turnover and new technologies for measurement of BMD have been developed that should greatly enhance the definition of fracture risk. To exploit these and evaluate their prognostic value and to develop and test models for describing bone loss and fractures in the population, we will enroll and follow prospectively a new random sample of Rochester women. 3) Although they account for a substantial proportion of age-related fractures, there have been no systematic studies on the causes of bone loss and fractures in men. Thus, we will enroll and follow prospectively a random sample of Rochester men to obtain this information and will compare and contrast results with those in women. 4) Bone mass is known to have strong hereditary component. We will search for genetic markers that might be used to predict future risk for osteoporosis by analyzing leucocyte DNA for restriction fragment length polymorphisms and skin fibroblasts for mutations in one of the genes for type 1 procollagen. 5) Although estrogen replacement therapy (ERT), begun at menopause, is the most effective way to prevent osteoporosis in women, it is not known whether it must be continued indefinitely for long-term benefit. We will attempt to resolve this important question by comparing values for BMD in a cohort of oophorectomized Rochester women, of whom some, but not others, received ERT for variable intervals. In all of these studies, we will continue to rely heavily on the Rochester Epidemiology Project, a unique resource for making population-based epidemiologic studies. We will employ state-of-the-art methods using dual energy x-ray absorptiometry to assess bone loss (including scans of the total and regional skeleton and lateral scans of the vertebrae) and to estimate lean body mass (for studying the relationship between muscle mass and bone mass), new biochemical markers (for assessing bone turnover), and molecular biology techniques (for assessing genetic predisposition to osteoporosis). This research will allow us to identify and quantify the determinants of bone loss and fractures in the general population so as to recognize better those individuals at risk for osteoporosis, who could then be targeted for preventive intervention.